[Gretl-users] Re: OLS regression with heteroscedasticy problem

Your main problem here is a lack of data. A "Breusch-Pagan test is nonsignificant (Chi-square(4) = 1.77291, p= 0.777434)" does not reject the null of heteroskedasticity. With your short series, however, the test has very low power. Check what you are doing here. Without some knowledge of the economic model that you are trying to estimate and your data, it is not possible to advise you. Perhaps you should check your application of the Breusch Pagan test to your pooled regression if that is what you have done. John C Frain 3 Aranleigh Park Rathfarnham Dublin 14 Ireland www.tcd.ie/Economics/staff/frainj/home.html https://jcfrain.wordpress.com/ https://jcfraincv19.wordpress.com/ mailto:frainj@tcd.ie mailto:frainj@gmail.com On Fri, 1 Apr 2022 at 13:39, 3J LEMA <3jlema(a)gmail.com&gt; wrote: